Membrane fusion is an essential process. Studies on viral and SNARE fusion catalysts have revealed a common strategy to overcoming the energy barrier to fusion, wherein favorable protein-folding reactions within complexes anchored in opposing membranes drive lipid bilayers together. However, a potentially new paradigm has arisen with discovery that atlastin (ATL), a membrane-anchored dynamin-related GTPase, can trigger fusion of synthetic liposomes, and is required for the branched morphology of the ER. ATL is unique because it is a mechanochemical enzyme that directly couples hydrolysis of GTP to fusion catalysis. Importantly there is not yet agreement on how this works. Is the initial membrane-tethering event upstream or downstream of GTP hydrolysis? Is nucleotide hydrolysis coupled to tethering, or fusion? Does the power stroke-like crossover conformation serve to tether membranes or to fuse them? What is the role of Pi release? This proposal is directed at answering these mechanistic questions. In aim 1 we have established a membrane-tethering assay, based on dynamic light scattering and cryo-electron microscopy, to measure tethering apart from fusion. This assay is being used to determine whether the initial tethering step requires GTP hydrolysis and/or crossover. Preliminary results indicate that tethering depends strictly on hydrolysis of GTP, ruling out a model in which a GTP-bound head contact mediates tethering upstream of hydrolysis. Whether crossover is also required for the tethering event is currently being tested. The result will help tease apart whether crossover contributes more to the tethering or the fusion step. In aim 2 we will use stopped-flow and quench-flow approaches to better understand the coupling between GTP hydrolysis, Pi release, crossover dimer formation and lipid mixing. Because much of our analysis will focus on ATL that is stably integrated into the lipid bilayer, the outcomes should give insights into conformational coupling for the full-length ATL protein in the context of the fusion reaction, thereby helping to distinguih amongst several contrasting models currently in the field. Finally, in aim 3 we will characterize a collection of known functional mutations in ATL1 that cause the motor neurological disease hereditary spastic paraplegia HSP. These will be assayed for in vitro fusion activity as well as for ER network forming activity in cells. Further characterization of mutants, defective in fusion per se, may give further insights into the fusion mechanism; whereas, characterization of mutants that can fuse membranes but yet cannot mediate network formation, may reveal new cellular regulatory mechanisms for ATL-catalyzed fusion. Because mutations in human ATL1 cause the motor neurological disorder HSP whose basis is not understood, these studies have the potential to shed light on disease causality and possibly also impact its therapeutics.